TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 1
TRD alignment with AliAlignmentTracks
Dariusz Miśkowiec, GSI Darmstadt
ALICE offline week, 10-Oct-2007
intro
optimizing alignment procedure
determination of residual resolution
summary
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 2
alignable objects in TRD
18 TRD supermodules like /TRD/sm03aligned by survey
540 TRD chambers like /TRD/sm03/st3/pl0aligned with tracks
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 3
alignment study via simulation: general idea
~100 k events pp:misaligned simulationideal reconstruction
alignment procedure
residual misalignment
Is it equal to the "canonical"residual misalignment used in PDC06 and PDC07?
Impact on efficiency &momentum resolution?
optimize!
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 4
alignment study via simulation: general idea
~100 k events pp:misaligned simulationideal reconstruction
alignment procedure
residual misalignment
Is it equal to the "canonical"residual misalignment used in PDC06 and PDC07?
Impact on efficiency &momentum resolution?
optimize!
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 5
alignment procedure with AliAignmentTracks
AliAlignmentTracks
AliTrackPoints.root
PointSelector.CAliESDs.root
AliESDfriends.root
AliAlignObj
AliTrackFitter (Rieman, ?)
AliTrackResiduals (Chi2, Fast,
Linear)
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 6
practicing the alignment procedure with AliAlignmentTracks
30 k pp events with ideal alignment (Silvia Masciocchi's production)
~ 300-400 tracks with pt > 0.8 GeV in each TRD chamber
pick one particular TRD chamber and align it to TPC
look at the residuals along phi
the peak should be at zero and as narrow as possible
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 7
AliAlignmentTracks with AliTrackResidualsChi2
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 8
AliAlignmentTracks with AliTrackResidualsFast
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 9
AliAlignmentTracks with AliTrackResidualsLinear
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 10
AliTrackResiduals daughters compared
time per ch
fixing params
ignoring outliers
working?
AliTrackResidualsChi2
34 s + – +AliTrackResidualsFast
3 s – – +AliTrackResidualsLinear
8 s + + –*
* setting fraction to 100% does not help changing σx from 100 to 1 cm does not help
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 11
external (to AliTrackResiduals) removal of outliers
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 12
Does it work for non-zero initial misalignment, too?
use 10 events with 200 particles with pt > 1 GeV
sent in the direction of one particular stack 400 tracks used to align one chamber
introduce misalignment at the beginning
of the alignment procedure
look at the residuals before alignment (just to get some feeling)
look at the residuals after alignment –
are they zero centered?
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 13
1 cm shift in phi – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 14
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 15
5 cm shift in z – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 16
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 17
1 cm shift in r – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 18
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 19
5 deg tilt in phi – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 20
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 21
5 deg tilt in z – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 22
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 23
0.2 deg tilt in r – before alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 24
... after alignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 25
alignment study via simulation: general idea
~100 k events pp:misaligned simulationideal reconstruction
alignment procedure
residual misalignment
Is it equal to the "canonical"residual misalignment used in PDC06 and PDC07?
Impact on efficiency &momentum resolution?
optimize!
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 26
resolution: results of running alignment on ideal data
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 27
... and how they compare to our "canonical" values
canonical values for PDC06:
rdphi dz dr rot phi rot z rot r (mm) (mm) (mm) (mr) (mr) (mr)------------------------------------------------------------------------ 3 3 3 0.4 2 0.4 supermodule initial 1 1 1 1 1 0.7 chamber initial 0.02 0.03 0.07 0.3 0.3 0.1 chamber residual------------------------------------------------------------------------
resolution observed when aligning 40 TRD chambers using
100 k events simulated with ideal geometry
0.09 0.70 0.25 0.8 0.8 0.2
factor 5 25 4 3 3 2
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 28
summary and next steps
AliAlignmentTracks with AliTrackResidualsFast works reasonably
residual misalignment probably somewhat larger than guessed
alignment resolution related to the width of the residual in many cases playing with one chamber is sufficient
under investigation
more optimization: pt cut, B field, scaling with statistics...
impact on efficiency and momentum resolution
related subjects
AliAlignObj storing local misalignment rather than global
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 29
backup
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 30
AliTRDalignment class – tool to manipulate TRD alignment sets
AliTRDalignment
ascii
root file
OCDB
geometry file
ascii
root file
OCDB
geometry file
converting between different file formats
generating random sets for simulation
reporting and visualization
survey file
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 31
What is being surveyed? Four survey points at each end ofeach supermodule. Unique names:
TRD_ sm08 a/c l/h 0/1
sm# z r phi
Survey file - ascii file in Alice-wide standard format
> Title:
ALICE - TRD Measurement of the first inserted TRD Supermodule 08
> Date:
14/12/2007
...
> Data:
TRD_sm08ah1 -3.6504 0.3337 3.5311 M 3
TRD_sm08ah0 -3.5451 0.9294 3.5306 M 3
processing survey data with AliTRDalignment
TRD_sm08al0
TRD_sm08ah1TRD_sm08ah0
TRD_sm08al1
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 32
AliTRDalignment a
create alignment object
a.ReadSurveyReport("Alice_TRD_5061.txt")
parse survey standard ascii file, decode and
store the survey x, y, z, error
a.SurveyToAlignment(8,"111000");
find such values of the 6 alignment parameters that, when applied to supermodule 8, minimize the chi-squared between the nominal and the measured positions of survey points.
processing survey data with AliTRDalignment
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 33
sm z r phi rphi z r
-------------------------------------------------------------------
local survey 8 0 1 0 -29.968 353.060 35.213
local ideal -30.250 351.000 37.450
difference 0.282 2.060 -2.237
local survey 8 0 1 1 30.525 353.110 35.239
local ideal 30.250 351.000 37.450
difference 0.275 2.110 -2.211
chi2 = 208.27
processing survey data with AliTRDalignment
phi: 0.3 cm off z: 2.1 cm off r: 2.2 cm off
comparison of Dec-2006 survey with nominal positions of survey points in local sm frame
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 34
processing survey data with AliTRDalignment
phi-shift z-shift r-shift phi-rot z-rot r-rot-------------------------------------------------------------- 0.279 2.085 -2.224 0.000 0.000 0.000 ±0.300 ± 0.300 ±0.300 ±0.000 ±0.000 ±0.000 (only translations)
0.295 2.085 -2.224 0.000 -0.025 0.000±0.477 ±0.300 ±0.300 ±0.000 ±0.568 ±0.000 (translations and z-rot)
phi: 0.3 cm z: 2.1 cm r: 2.2 cm
to larger phi i.e. away from inward
downward muon arm
alignment params of supermodule 08 deduced from Dec-2006 survey
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 35
processing survey data with AliTRDalignment
shift in z – caused by the survey target offset
nominal positions of the survey points in AliTRDalignment modified to account for the survey target offset of 20 mm
TRD alignment with AliAlignmentTracks, D. Miskowiec, Alice offline week, 10-Oct-2007 36
processing survey data with AliTRDalignment
779
"r"-coordinate of the two surveyed points:365.25 cm from survey365.75 cm from drawings (Bernd)367.10 cm from offline (my guess, based on BTRD being 779 mm thick)
775
needs to be understood